ESPE Yearbook of Paediatric Endocrinology

JAMA. 2022;327(4):350-9. doi: 10.1001/jama.2021.23686.PubMed ID: 35076666

Brief summary: This study assessed the penetrance of pathogenic and loss-of-function clinical variants in 2 large population-based biobank studies. It found that the penetrance of pathogenic/loss-of-function variants was variable but generally low among such unselected populations.

We are increasingly using genetic tests to confirm the diagnosis in patients whom we suspect may have a (often rare) monogenic disease. Such tests use DNA sequencing of the patient to detect those who have pathogenic and loss-of-function clinical variants in known disease causing genes. However, as genetic sequencing is becoming more popular, for example as a screening test in ill patients or even as a predictive tool used by healthy individuals, it is becoming clear that the diagnostic value of a positive test result is much weaker than we are used to. This is because, in our usual clinical practice, we highly select patients for genetic tests based on their clinical presentation; e.g. we select patients with early onset T2D and positive family history for MODY (Maturity-Onset Diabetes in Youth) for gene sequencing – and all patients who return a positive genetic test are given the diagnosis of MODY. By contrast, if we perform the same test in 100 000 unselected adults, only 20–30% of those with ‘pathogenic’ mutations in HNF1A and HNF4A actually have diabetes (assessed by HbA1c testing and/or clinical records) (1).

The current study used the same approach across a wide range of genes for 197 diverse monogenic disorders in large biobank studies, the UK Biobank and the BioMe Biobank (Mount Sinai Health System in Manhattan, New York), who underwent whole exome sequencing. There was very wide variation in the penetrance of known pathogenic mutations. Even for well-established gene-disease pairs, the penetrance of pathogenic/loss-of-function variants varied enormously: in LDLR for familial hypercholesterolemia (penetrance 0%–100%); in BRCA1, BRCA2, and PALB2 for breast cancer (0%–100%).

These consistent findings challenge our notion that pathogenic mutations for monogenic disease should segregate highly (almost perfectly) with disease status. Instead, the prior reasons for choosing a test are crucial for the interpretation of the result. Furthermore, the predictive utility of genomic testing as a screening tool in wider populations requires further consideration.

Reference: 1. Mirshahi UL, et al. The penetrance of age-related monogenic disease depends on ascertainment context. 2021; medRxiv: doi: https://doi.org/10.1101/2021.06.28.21259641.